Beyond human: Non-human minds elicit both positive and negative implicit evaluations

Interactions with non-human “minds” are now ubiquitous. What are our implicit evaluations of such entities like? Past work has found that specific persons elicit bivalent priming, suggesting the coactivation of positive and negative implicit evaluations. This contrasts with common attitude objects that elicit univalent priming, suggesting the activation of either positive or negative implicit evaluations. Is implicit bivalence triggered only by specific persons or might it also be triggered by non-human minds, such as a social robot? To test the hypothesis that implicit bivalence emerges even for non-human minds, across four studies (total N = 349), we led participants to believe that they would interact with a robot that had the mental capacities of a three-year-old child, and as a within-person comparison, another ordinary toy (e.g., a toy car). Prior to any interaction, participants completed an evaluative priming task to assess implicit evaluations. Priming participants with the name of the robot facilitated the classification of both positive and negative targets (bivalent-priming). In contrast, priming participants with the name of the comparison toy only facilitated the classification of positive targets but not negative targets (univalent-priming) or had no significant effects. This work provides robust evidence that implicit bivalence, previously demonstrated only in relation to specific persons, can be elicited by non-human minds (i.e., social robots). Findings suggest that perceivers create, “on the fly,” a representation of a mind, even a non-human one, that is linked with positive and negative. We discuss implications for implicit bivalence and implicit evaluations more broadly.